OK, before I say this, let me say that I have only a little bit of WEAK evidence to say this theory is possible. It is ONLY an idea I've had for years, but have been unable to test. With all of the unknown hybrids out there, it may be impossible to test, too. Anyway, I'm going to go out on a limb here and say my theory, though.
Assume that the pattern mutation in southern pinesnakes that results in the patternless trait is at the same loci in southern pines as the blotched trait in northern pinesnakes. Now, let's say that the blotched trait allele is dominant to the patternless trait allele in southern pines. HOWEVER, let us say that the patternless gene is dominant to the northern pine allele for the normal blotched pattern. What this means is that a het PURE southern pine would be normally patterned, but a het F1 northern pine hybrid would HAVE to be patternless.
Once the hybrids get all confused following generations of crosses, you could have an animal that is believed to be a pure southern that really has one or more northern pine alleles which means that sometimes iut acts dominant and sometimes it acts recessive depending on if the snake in question has which genes.
Possible, but unlikely. Still, it answers all of the data I've ever heard of, so it can't be disproven. The only stumbling block to me is that I find it hard to believe that the patterns of two closely related subspecies, that intergrade in the wild, would be inherited so separately.
Anyway, I'm out on the limb now.....lol.
KJ